Fluorometry LG Chatten MANIK

Chapter florometry

1. CHAPTER 4
Fluorometry
David E. Guttman
SCHOOl. OF PHARMACY
STATE UNIVETY OF NEW YORK AT BUFFALO
BUFFALO NEW YORK
4.1 Introduction ..........................167
4.2 Theory ............................168
4.3 Fluorescence and Chemical Structure ...............169
4.4 Instrumentation for fluorometry .................173
4.5 Factors Influencing Intensity of fluorescence ............175
A.Concentration of fluorescing Species ..............175
B.Presence of Other Solutes ..................181
. Hydrogen-Ion Concentration .................183
D.Temperature ........................187
E.Other Factors .......................188
4.6 Comparisons of Fluoromeiry with Spectrophotometry ........188
A.Sensitivity .........................1 88
B.Specificity .........................189
C Experimental Variables ...................I 89
4.7 Application of Fluoromctry to Pharmaceutical Analysis ........90

4.8 Practical Section .........................197
A.General ..........................197
B.Laboratory Projects in Fliioromcrry ..............198
Problems ..............................200
References .............................* 201
4.1 INTRODUCTION
The mechanisms by which molecules absorb cicctromaonetic radiation in
the visible and uliras,olct rernons of the spcctrum and are, as a result, raised
167

2. 168 I LCOKQMJ TRY fcn. 4)
to excited ciccirnic states were discussed in Chapter I. The reverse process,
loss of enerev and conCOmlijuir transition of moleculcs from excited states
to ground states of cncry. can occur with recmissuon of radiatidn. Such
emission is known a' Iu,ni,wsc,',ic.,1hc Intensity and compositio'n of the
-emitted radiation can he nicsurcd anduch measurements Iorm the basis of
sensitive method of analysis called _11ijoroineir
.
it Fluorometric methods of
analysis have found apphcaiitiri In many situations' of pharmaceutical interest
such as in the anaksis of riboflavin. thi;iniiric. and rcccrpunc in drug dosage
forms. More sueniticant and %kidespread. huscer. has been the application
of flunronictrrc iechniqucs in the analsi of trace amounts oIdrugs and meta-
bolites in biological tissues and fluids.
4.1 THEORY
"Absorption of ultraviolet and visible light by molecules of an irradiated
sample generates a population of molecules in excited electronic states. Each
excited electronic state has many different vibrational energy levels, and
excited 'molecules will be distributed in the various vibrational energy levels
of an excited state. Most usually this state is a thzglei store, i.e., one in which
all of the electrons are paired and in each pair the two electrons spin about
their own axes in opposite directions. It is intuitively apparent that since the
amount of radiation absorbed by a sample does not decrease if adiation is
continued, efficient rapid processes must be operant which result in the loss of
energy by excited molecules and their return to the ground state. It has been
calculated, in fact, that the average lifetime of molecule in the singlet excited
state is of the order of 10-9 sec Molecules at each vibrational level of the
excited state could, for example, lose energy by emitting photons'and as a
result fall to the original condition of the ground state. The energy and, there-
fore, the wavelength of emitted light would then be cactIy the, same as that
absorbed. Such a process is termed resonance fluorescence) It is an improb-
able process and is rarely encountered in solution chemistry. Rather, mole-
cules initially undergo a more rapid process, a radiationless loss olvibrational
energ. and so quickly fall to the lowest vibrational energy level of the excited
state. tThe vibrational.cnergy is thought to be lost to solvent molecules. The
process is knoxn as vibrational level
of the excited stale .rnolccji lQ.can ciLhcz xcWrn1a LhcgrQI state by photo-
emission or by radiationkss processes. If indeed the former occurs, the emis-
sion is a type of luminescence referred to asfluorescrnre. Fluorescence is
dcfincd as the radiation emitted in the transition of a molecule Tm"iTsiiiet
excited state toangIct ground state. Because ofvibratinl iLlaxation in the
excited slate-and because" miilcciilc may return to a vibrational level in the
ground state which is higher than that initially occupied prior to excitation.
the radi.iiuon emitted as fluorcscencc is of loter cnerizy and, therefore. of
longer cavclength than that orirnallv absorbed.

3. - 4.3 FLUORESCENCE AND CHEMICAL STRUCTURE 169
Other processes involving the excited state can occur to compete with
fluorescence emission, and not all of the absorbed energy will be emitted as
fluorescence. The extent to which such other processes occur is characterized
by a parameter known as the quantum efficiency of fluorescence which is
symbolized by and defined as the ratio of the number of light quanta emitted
to the number absorbed. Quantum efficiency approaches 1 for highly fluor-
escent compounds and 0 for those which fluoresce weakly. It is interesting and
pertinent to consider the processes which occur to decrease the efficiency
of fluorescence(An excited molecule could, for example, undergo a radiation-
less loss ofcnersufficient to drop to the ground state. This process is termed
internal conversion. With some compounds a process known as intersystem
ngnobcur. Here a molecule in the lowest vibrational level of the
excited state converts to a-.ipkEi#àr a state lying at an energy level irter-
mediate between ground and excited states and characterized by an unpa4ixig
of two electrons. Thus, in contrast to the singlet state, there is a spin re'tersal
involving one electron of a pair and the two electrons spin about their axes in,
the same direction. Once intersystem crossing has occurred, a molecule
quickly drops to the lowest vibrational level of the triplet state by vibratiqnal
relaxation. The triplet state is much longer lived than the corresponding
singlet state with lifetimes of lO to 10 sec. From the triplet state a molecule
can drop to the ground state by emission of radiation. This type of lumines-
cence is termed pjasp)wretccrzte and is formally defined as emission of radia-
tion resulting from the transition of a molecule from a triplet excited state to a
singlet ground stai! Phosphorescence is often characterized by an afterglow,
i.e., because of the long life of the triplet state, luminescence can be observed
after the source of exciting radiation has been removed. In contrast, no
afterglow is observed in fluorescing systems because of the short life of the
singlet excited state. A molecule in the triplet state can also undergo radia-
tionless conversion to the ground state. Such a conversion is enhanced by the
relatively long life of the triplet state so that collisions fruitful in dissipating
energy can occur and by the fact that the energy difference between triplet
state and ground state is not inordinately large. These processes are dia-
grammatically illustrated by Fig. 4.1.
4.3 FLUORESCENCE AND CHEMICAL STRUCTURE
Quantitative aspects of the processes which can in-volve the excited 'elec-
tronic state arc not sufficiently well understood to permit predictions as to
whether or -not a particular compound will fluoresce to a degree necessary for
analytical purposes.(,Fluorometric methods are, of course, limited to those
compounds which possess a system of conjugated double bonds. A com-
pound must absorbradiaton in order to fluoresce, and it is the plLsence within
a molecule of the mobile 77 electrons which is responsible for absorption

4. E .c iid
singlet trot.
Ground stole
Eicired
trpICt 11011
170 I I I:Oir)%Il TRY ic.
1r,ILterllIc ii the iohIc and ultraviolet reclorh I the spectrum. The
prceilCe i'I :1 hr.'in'piiorc" does not occcssarilY endow -I 'ith
the abihcv to Ilitiresce since radiatuisniess processes and tnrcrsvstcm dossun
can occur to decre:be the quantum efficiency ol fluorescence to zero or to a
dcercc that makes practical measurement iitposthlc. It sould be expccted,
h.cvcr. that structural features which influence the dceree o[ conjugation of
a molecule and the dcloealizauon of irelectrons mittht iritlucncc the likelihood
FIGURE 4.1: A diagrammatic representation 01 the changes in energy levels 01 a molecule
that can occur as a result of the absorption of electromagnetic radiation: (I) absorption of
radiant cncry boosting molecules to various vibrational energy levels of the excited singlet
state: (2) radialtonlcss vibratiorui relaxation to the lowest vibrational level of the excited
singlet state; (3) rodiatiorikas internal conversion from excited singlet state to ground state
followed by vibrational relaxation; (4) fluorescence followed by vibrational relaxation;
(5) intcrsystem crossing from excited singid Mate to excited triplet state; (6) vibrational
relaxation to the lowest vibrational level of the excited triplet state; (7) radiationlcss inter-
nal conversion from excited tripld state to ground state followed by vibrational relaxation;
(8) phosphorescicc followed by vibrational relaxation.
of measurable fluorescence. Thus, for example, saturated compounds such as
cyclohcxane are nonfluorescent, while benzene is weakly fluorescent, and
highly unsaturated polycvdic aromatic compounds such as anthracene are
strongly fluorescent. Similarly, the reduced form of riboflavin (I) does not
have the degree of conjugation of the parent compound (II) and does not
possess the fluorescence characteristics of riboflavin.
C}l,—ICHOI tl,—C1-1,OH CH,—(CHOH)1—CH2OH
I t CII
CH, (4NO
Cif
.1Y_tl
CII
,
U
MI

]ff
Definite correlations between chemical structure and lluureccticc.cannot be
made. However, the work and review of \Villianss and Bridges' does provide
ornc 1115i0it to the rnfl ucr.ec of structural features on the fluorescence of

5. 43 FLUORESCENCE AND CHEMICAL STRUCTURE 171
organic compounds. Monosubstitution of benzene with alkyl groupings, for
example, was found to have little influence on the intensity of fluorescence of
the substituted benzene relative to that of benzene. However, monosubstitu-
don with groups known to increase electron delocalization (ortho.para-
directing groups), such as fluoro, amino, alkylaminodialkylarnino, hydroxy,
and methoxy, yielded compounds that fluoresced more intensely than did the
parent compound. Substitution with iodine, chlorine, and bromine resulted
in benzene derivatives which either did not fluoresce or fluoresced to a lesser
degree than benzene in spite of the fact that these substituents arc also ortho-
para directing. This influence of halogen substitution was also reported by
McClure' and is apparently due to an enhanced intersystem crossing process
since bromine- and iodine-substituted aromatic compounds exhibit intense
phosphorescence but only weak fluorescence. Most meta-directing substitu-
ents (which tend to localize ir electrons) were found to markedly decrease
fluorescent intensity. Thus, benzoic acid, nitrobenzene, bcnzenesulfonic acid,
benzcnesulfonamide, and benzaldehydc were found to be nonfluorescent:
Bcnzonitrile, in contrast, fluoresced more intensely than benzene, even though
the CN group is meta-directing. It was postulated that electrons of
the CN group interacted with the ir electrons of the benzene ring to result
in a distribution that favored fluorescence.
The fluorescencecharacteristics of disubstituted benzenes were also studied,
but few generalities could be generated from the results. Observed effects
were not predictable and apparently were the result of a combined influence
on the mobility of 7r electrons. For example, it might be expected that sub-
stitution of the fluorescent compound, aniline, with a meta-directing group
such as —SO,NH, would result in a compound which would fluoresce to a
lesser degree than aniline. Sulfanilamide, however, was found to be five times
as fluorescent as aniline. -Similarly, guides to predicting the behavior of
heterocyclic compounds could not be made because of the uncertainty of the
substittient effect. In general, it was found that a doubly bonded nitrogen
(=N—) in a ring tended to decrease the likelihood of fluorescence, while the
presence of —NH—, —.0---, and —5-- appeared to contribute to the likeli-
hood of fluorescence.
Molecular geometry must also be considered in attempting to relate chemi-
cal structure and fluorescence. That geometric considerations are important
is well illustrated by examples cited by "ihryand Rogers. Fluorescein (Ill),
for example, is highly fluorescent, while p. nolphthalein (IV) is nonfluores-
cent. The oxygen bridge in fluorescein imparts ri the molecule a rigidity and
o:i-oo
oo-
yc^
(III) (IV)

6. 72 ii (iii'.iit,
FLu. 41
pf.iii:irits thu i :t prccnI in pheinulphih.ilcuru. I lie iDCflCe Lit a planar.,
ruid sirructUic [)eMiut' ihraiions and rOtliofl', ol itic aromatic rifl Co occur
.f1Li rcLJi( iii tidu.Iii)flies niissupCior i c\LiI.u1iii cuierv. SurtuuIar.• .uili
luururer 1 ci.irpiurid that ehirri nv-truly in:ricr1nm. the (1$ isomer
tIiivrece rrricui less intensely than the trans. Sit lhciie I V) is such a compound.
This can also be ascribed to a planaror cIhcins ith the cis is being nwi-
rl:1r1:tr due to the bulkiness of the :irinrn.Itre rr:rs
.,,—i II
1k—:
I'
That a compound does not 1uorcsce or has a lo%k iii flsitY of fluorescence
does not flCcCS5JrIly dismiss tluoronictrv as a potential tool for the analytical
determination of that compound. Many seil-acccptcd and widely used
fluorometric procedures are based on chemical conversions of weakly fluores-
cing compounds to derivatives which fluoresce intensely. For example.
tetracycline (VI) has a s%cak native fluorescence, but complexes of the anti-
biotic s oh Ca and a barbiturate fluoresce quite intensely.' Corticosteroids
N(CH3)
014 1
al,
OH
OH 0
CONIi.
ivi)
such as hydrocortusonc (VII) do not fluoresce. However, they form, in
concentrated sulfuric acid and in the presence of ethanol. strongly fluorescing
compounds.5 N-MeIhylnicotinamde (VIII) is determined in biological fluids
C}{O El
HO
by a I)uoronietruc method even uhoueh it hashutilc pause fluorescence. Here
the amble is condensed uth acetone and treated s'Jtji ha-bc Co vied fluorescent
IJ
_CONhhi
cl
Lviii

7. 4.4 INSTRUMENTATION FOR FLUOROMETRY 173
products. Similarly, epinephrine (IX) is assayed fluorometrically by measur-
ing the fluorescence of products resulting from oxidation and hydroxylation.'
HO——H—CH1—NH--CH3
HO
H
(IX)
4.4 INSTRUMENTATION FOR FLUOROMETRY
In contrast to spectrophotometry, the intensity of light transmitted by a
sample is not of direct concern in fluorometxy. Rather, it is the intensity of
radiation that is emitted as fluorescence that is measured and related to the.
tzdtotion Uonoctwom.tsr
or Fill.:
Sompti Hold.,
Emission Uonochromster
Of Filter
I kwiiiiV
Light Souvc.
Detector
LLL!J
Motor or Recorder
Amplifier
FIGURE 4.2: A diagrammatic representation of an instrument used to measure intensity
of fluorescence.
concentration of fluorescing species. The components of instruments which
are used in fluorometsy are, however, quite similar in design and function to
those employed in spectrophotometers and colorimeters. A diagrammatic.
representation of such a device is shown in Fig. 4.2.
The chief components are: a source of exciting radiation, an excitation
filter or monochromator by which a band of exciting light can be isolated to
be passed on to the sample, a sample holder, an emission filter or mono-
chromator by which a band of fluorescence can be selected for detection, a

8. 174 I I.L'OROMETRY cii. 4)
detector, and some means for amplifying ;Ind indicating the detector response.
In most commercially available instruments. the detector is placed ac a right
angle to the direction of travel of the beam of exciting light. This arragcment
hasbccn found to be the most vdvantapcous for mcasurinhe fluorescence
f dilute solutions. Other arrangements arc possible. however. since fluorcs-
cerlcc is emitted in all directions. Thc'liiht source is usually a mercury or
cnon arc. Those instruments shich use litters as coarse monochromcters
.rc cierred to asJfuorurnctcrs. Those employing more sophisticated and cact
t!rat:ng or prism monochromators arc termed
spectrometers, or spcctrup/iotofluvrotnewrs. The first term appears to be
the one most commonly used to designate this type of instrument. The other
components of an instrument such as sample holders. detectors. amplifying
and indicating devices arc much the same as those discussed under spectro-
ph a tome try.
Spectrofluorometers arc used in tluorornctric work in a manner analogous
to the use of spcctrophotomcters in absorption spectrophotometrY. They
enable an investigator to gcneratc two types of spectra which arc pertinent to
fluoromcry. The excitation spectrum is obtained by setting the emission
monochromator at a suitable wavelength and measuring the intensity of
fluorescence as a function of the wavelength of the exciting radiation. In
thcor, the maxima and minima exhibited by the excitation spectrum should
be at wavelengths which arc identical to those found in the absorption spec-
trum of the compound. In practice, exact coincidence may not be found due
to instrumental artifacts. The emission spectrum of a compound is obtained
by setting the excitation monochrometer at an appropriate wavelength corre-
sponding to strong excitation and measuring ihc intensity of fluorescence as a
function of the wavelength of emitted light. An example of excitation and
emission spectra is shown in Fig. 4.3 for grisoefulvin in 1 % ethanol. As
would be expected from the considerations discussed in the theory section,
the emission spectrum is found at longer wavelengths than the excitation
spectrum. Some overlap of the two spectra is frequently observed.
Fluorometers are used in a manner somewhat analogous to calorimeters in
absorption work. The excitation and emission spectra of compound dictate
the transmittance characteristics of filters that should be employed for a
particular analytical determination with a fluorornetCr. The filters should be
as much as possible mutually exclusive. That is, the emission filter should not
pass wavelengths which arc transmitted by the excitation filter. This preau-
Lion is necessary to preclude interferences from tight whichmay be reflected
by the sample holder and other parts of the instrument and frm light scattered
by the solvent uscd l:luoromcters arc somewhat more sensitise than spectra-
tluoromctcrs since filters pass a more intense radiation than prism or grating
monochromators. Fluorometers are recommended for routine quan(itZItiC
work, while spcctrofluarumctcrs arc necessary research tools in tlic develop-
ment of fluorometric assay methods.

9. 4.5
FACTORS INFLUENCING INTENSITY OF FLUORESCENCE 175
A variety of
fluorometerS and spectrofluorometers are available from manu-
facturers of scientific equipment. Figure 4.4 is a schematic diagram of the
optical system of a widely used filter flu orometer (model 110, C. K. Turner
Associates). Figures
4.5 and 4.6 illustrate the appearance and optical charac-
terisXis
of a spectrofluorometet (Aminco-Bowman spectrophotofiuorometer.
100
EICIIOtSo4 Spectrum
fsmiteioA monothrosn,lef)
iii at 450 MpA
Em.suon Spectrum
(ec1IQI0fl moeoctwemelsrl
sit 01 295 MM
"--'i-I
I,
eo
..
U
V
—
=f 40
0
C
20
0
250 . 350 450 550
Wavelength. m
FIGURE 4.3: Excitation and emission spectra of griseofulvin in water containing I
ethanol. Reprinted from Ref. 41. p.
365, by courtesy or Nature.
American Instruments Company. Inc.). The characteristics and features of
manyavailable instruments have been the subject of recent excellent reviews.
It is recommended that the reader consult such reviews and the literature
available from manufacturers for specific information on commercially
available instruments.
4.5 FACTORS INFLUENCING INTENSITY OF
FLUORESCENCE
A. CONCENTRATION OF FLUORESCING SPECIES
The relationship between observed intensity of fluorescence and the con-
centration or fluorescing species is considerably more complex than that

10. MOTOR
FILTER
P.a.ee oiy uv
76 FLUOR0Mr:rRy (cli. 4
FIGURE 4.4: A schematic diagram of the optics of the Turner, model 110 fluoromcter.
Courtcsy of G. K. Turner As.socialcs.
betccn absorbance and the concentration of absorbing species which is
dictated by Beer's lawJ Complexities arise from both theoretical considera-
tions and from practical aspects of the instrumentation used to measure
fluorescence. It might be intuitively anticipated thatjaarctonhip.
might not exist since fluorescence intensity would be expected to be propor-
ticwaLitoa the coQn olmoiccuks in thrxcjced tae and, therefore.
proportional to the intensity of radiation responsible for excitation However,
FIGURE 4.5: The Aminco-flowman 3pctrophotofluoromcIcr. Courtesy of the American
Instrument Company. Inc.

11. 3 5 FACTORS INFLUENCING FT:NSITY OF FLL0RESCECE 177
FIGURE 4.6: A
schematic diagr2m of [he optics of the Amico-Bo.man spectrophoto-
fIuoromctcr. Courtesy of the American Instrument Company, Inc.
light is absorbed by the sample and (he intensity of exciting li
ght does not
remain constant, but diminishes as the light beam traverses a sample. If
the Solution is dilute, stintIicartt absorption will not occur and the decrease in
tntcnsity of exciting light will not be sig
nilicarit. In a more concentrated
SOlUtIOn the intensity of exciting light might well be different in different
regions or the sample
being irradiated. This consideration as formalized by

12. 178 FLUOROMETRY [CH. 4]
Kavanagh" and in more detail by Braunsberg and Osborn," who reasoned
that fluorescence intensity should be proportional to the amount of light
absorbed by the sample, i.e.,
F=k#(10—I) (4.1)
where F is the intensity of fluorescence, k is the proportionality constant,
0 is the quantum efficiency of fluorescence, 1. isthe intensity of light incident
on a sample, and I is the intensity of light transmitted by a sample. Since, by
Beer's law,
I =
1e' (4.2)
where £ is the molar absorptivity of the compound at the wavelength of'the
exciting light, b is the path length along the axis of irradiation, and c is the
concentration in moles per liter, Eq. (4.1) can be rewritten:
F = kç610(1 - e') (4.3)
The term in parentheses can be represented by a series expansion
I
- 2! n!
to yield:
bc (bc)2 (bc) 1

F = kc6I.bc[1 - -- + - +
(n + 1)!]
(4.4)
The detector of a fluorometer does not measure total fluorescence intensity.
but rather the intensity from only a segment of the sample. Equation (4.4)
must then be modificd
cbc (bc)2 (bc)'1
Sr = kccg8!oEbc[l - + -
+ (n + l)!j
(4.5)
where S, is the electric signal generated by the detector, g is the constant re-
flecting the sensitivity of the detector and the amplification of the detected
signal, and 0 is the constant reflecting the geometry or the system, particuIrly
the solid angle of light viewed by the detector. As recently discussed by Her-
cules,l2 there arc two concentration regions where Eq. (4.5) can be con-
veniently simplified. When the concentration c is very small, the equation
can be approximated by:
S, = kcg0J0bc (4.6)
Such an approximation is reasonably valid when £bc < 0.05. Under such
conditions, a linear relationship between measured intensity of fluorescence
and concentration exists. When the concentration is large, e—bl approaches
zero and Eq. (4.5) can be approximated by:
S, = kg010 (4.7)
Udder this concentration condition, measured intensity is independent of

13. V 1.0 2.0 3.0
80
C
2 40
22
U
C
20
0
-p
4 5 FACTORS INFLUENCING I.TESITy OF FLUORESCENCE 179
I
concentrationj At concentrations intermediate to the to extremes, a non-
linear relationship would be theoretically expected. It is interesting and rcie-
vcnt to note that the Concentration range over which linearity is theoretically
expected is dependent on the molar absorptivirv of the compound. This i
illustrated in Fig. 4.7, which shos the types of intensityconcentrationpro
tiles which can be theoretically expected for compounds having different
absorptivities. It can be seen thaover the concentration range of the graph.
Motrit1 x
FIGURE 4.7: Intenslty'conccn(ra(jon profiles for three different solutes haing different
molar absorptivicics. Reprinted from Ref. 12. p.31 A. by courtesy of Analvtkol Chemistrv.
a linear relationship holds for the compound with the 10%
,,cst absorptivity.w
hile the compound with a high absorpivitv exhibits a linear relationship
o'er a rather small range of concentrations.
A further complication arises if the excitation and emission spectra of the
Compound overlap. In such a case, photons emitted as fluorescence can be
absorbed in exciting other molecules and will not be measured as fluorescence.
In dilute solution such an occurrence sill probably not be significant.
However, in concentrated solucion, Kit-absorption of fluorescence radiation
can occur and ijl result in a mcaured intensitv'.hich is less than that pre-
dictable on the basis of E. (45).

14. 100 200 300 400
so
so0
40
•0
'-a
•0
C-.
. zo
180 FLUOROMETRY (c"
4)
A more serious problem which is concentration-related can result due to
the geometry of the measuring system. A detector set to view fluorescence
at a 90° angle only a small band in the center of the sample. With dilute
solutions, this band emits fluorescence which is representative of the whole
cell. However, if the sample is concentrated, sufficient light absorption
might occur so that the portion sensed by the detector is only weakly irradi-
ated. This results in the phenomenon of concentration retersal. i.e., an
Con tri$rati., Meg/ml
FIGURE 4.8: A plot showing the influence of concentration on the intensity of fluorescence
of solutions of phenol. The excitation wavelength was 295 nip and the emission wavelength
was 330 nip. Reprinted from Ref. 1. P. 377. by courtesy of the Journal Of
Climic6l Path0h'gy
by permission of the authors, the editor, and the publishers. B.M.A. House. Tavistock
Square, London W.C.I.
increase in concentration results in a decrease in measured fluorescence in-
tensity. Such behavior is illustrated in Fig. 48 for phenol.
For these reasons, fluorometric measurehents made for assay purposes
are restricted to dilute solutions where a linear standard curve can be obtained.
In general, a linear response can be expected for solutions which absorb less
than 5 %'of the exciting radiation. It is apparent from Eq. (4.) that adecrease
in measured fluorescence resulting from a reduction in concentration can be
compensated for by an increase in the intensity of exciting radiation and/or
an increase in the sensitivity of the detector. Because of this, accurate meas-
urcjncnts can be made on relatively dilute solutions and a wide range of

15. 4.5 rACTOLS INFLUENCING INTES1TY OF FLUORESCENCE
concentrations can be covered in the linear portion of the intensity-concert-
(ration profile. This is in marked contrast to absorption spectrophotometry.
where a limited.range of concentration is necessary for accurate measure-
ments.
- \..-PRESENCE OF OTHER SOLUTES
The presence in a sample of soluts other than the solute being analytically
determined can influence the Intensity of fluorescence by one or more of a
number of different effects.
l.4luorescent Impurities
An obvious possibility is that another component of the solution might also
fluoresce and thus interfere with the determination. Impurities introduced
into the sample from solvents, buffers, detergents which are residual on
glassware, and from the source of the sample can fluoresce and can introduce
error. This possibility should always be anticipated in fluorometric work,
particularly if measurements are made on very dilute solutions, and especially
if the excitation radiation is in the ultraviolet region of the spectrum. Appro-
priate pi-ecautions must be taken which include the use of pure solvents and
chemical reagents and cleanliness in all operations.
2. lnler-F;lter Effect
The presence in solution of other solutes which are nonfluorescenc can
effect fluorescence intensity by the so-called inner-filter effect. The influence
here is due to the absorption of light and is similar to the situation discussed
previously where excitation and emission spectra overlap. Thus, if the non-
fluorescent components absorb either excitation or emission radiation, a
reduction in measured intensity of fluorescence will result. HerculesU has
considered some theoretical aspects of the inner-filter effect and has con-
cluded that if absorption due to other species is constant and if absorption
due to the fluorescing species is 'small, then a linear relationship between
measured fluorescence and concentration should still be observed. In such a
case, Eq. (4.6) assumes the following form:
5, = kdg0I0thc(K) (4.8)
where K is the constant resulting from the presence of other absorbers in the
system. It is apparent that if this effect is encountered the concentration of
the nonnuorcsccnt absorber must be eliminated or be maintained constant
from sample to sample and a standard curve must be used which was deter-
mined at that concentration of absorber. An alternative might be to charge
the wavelength of excitation or emission radiation to minimize this effect.
3. Chemical Quenching
In
addition to the cffcts just discussed, dissolved solutes can result in
decreased fluorescence by at least two types of chemical quenching processes.

16. 182 FLUOROMETRY
[cn. 4]
ne is known as collisional quenchin and results from a diffusion-controlled
process in which a
rnolccu c o quencher" interacts with an excited molecule
of the potentially fluorescing substance. Interaction results in the dissipation
of excitation energy not by fluorescence but by transfer of energy to the
quenching molecule.. A simplified mechanism can be written to describe this
situation:
F +J'r,.-.-.- F'
k
F. F + hr,
F' + Q F + Q
Here, excitation radiation h,, converts the "fluorophor" F to the excited
state. The excited molecule F' can dissipate excitation energy by fluorescence
hv, or by interaction with the quencher molecule Q. The rate constants k1, Ic5,
and/c3
characterize the rates of the various processes. The intensity of fluo-
rescence will be proportional to the steady-state concentration of molecules in
the excited state. In the absence of quenching agent, this concentration is
given by: - -

(F')0
=(
F) (4.9)
In the presence of quencher, the concentration of excited molecules will be
reduced and is given by:

(F')'
= k1(F)
(4.10)
k2 + k3(Q)
The ratio of intensities in the absence (f) of quencher to that in the presence
(f,) is, therefore,
JJJ,- I +k3/k2(Q) (4.11)
Equation (4.11) is known as the Stern-Volmer law and predicts that the ratio
of the two intensities will be linearly dependent on the concentration of quen-
ching agent.
The molecular basis for dissipation of energy by collisional quenching is
poorly understood. A recent hypothesis was discussed by Hercules' and
involves electron transfer between excited state and quenching agent. It may
be represented as follows
F' + Q - F :Q
/
1F+Q F + Q
The excited molecule F*
interacts with quencher
Q, abstracting an electron to
form an ion pair F:
Q-. The ion pair can dissociate to give either a triplet
state IFand
Q or a ground slate Fand Q. Both processes result in thermal
dissipation of energy.

17. 4.5 FACTORS INFLeE?JCNG I.fTY OF FLUORESCENCE :183
Another type of quenching is called static quenching. Here, complex forma-
tion occurs between a potentially fluicing molecule in the ground state
and a quencher molecule. If the complexed form of the potentially fluores-
cing molecule has different spectral characteristics than the free form. it
mamay not undergo excitation or m;_be excited to a lesser degree that the non-
complexcd species. Suppose, forexampk, that complex formation occurred
and resulted in the formation of acomplex, with a stability constant of K.
that was not fluorescent. i.e.,
F + 0 F:Q
F + /v.,—+ F'
k

F'—o-F+hr, -
The steady-state concentration of F* in the absence of Q is given by

(F')0 = kJk5(F), (4.12)
where (F), is the total concentration of potentially fluorescing compound.
In the presence of quencher,
(P).

(4.13)
'1+K,(Q)k2
The ratio of fluorescence in the absence to that in the presence of static
quencher is then given by: -
fJf l+ K1(Q) --- (4A4)
The reduction in fluorescence caused by this type of quenching is thus dic-
tated by the stability constant of, the complex and by the concentration of
quenching agent. - - ------.-.
Many examples of chemical quenching are found in the literature. Halide
ions such as iodide and chloride are well known examples of collisional
quenchers. Caffeine, related xanthines• and purines" have been shown to
influence the fluorescence of riboflavin by static mechanisms. In the usual
case, quenching is an undesirable effect and the possibility of encountering
this type of interference should always be evaluated in developing a fluoro-
metric assay. It is possible to utilize this phenomenon, however, as an analyt-
ical means for determining the concentrations of compounds known to
quench fluorescence. In addition, it is apparent from Eq. (4.14) that it is
possible to employ fluorometric methods for the investigation of complex-
forming equilibria.
- C. HYDROGEN-ION CONCENTRATION
The intensity of fluorescence..emitred by sgbiiinnso! 'vcak acid.sa.ndwgk
ses.n.e.xhibjt a dependtncy an the pH of the solutionIhe effect here

18. 0
'C
4&
C
0
?.1.
0$1
C
0
C
043
S.
p.
0
S
184 FLUOROMETRY
[CH. 4]
maybe due simply to a change in the degree of ionization of the weak electro-
lyte- For example. suppose that the fluorescence of a weak acid HA was
investigated in dilute solution using an emission filter or monochromator
- -.
a
PM
FIGURE 4.9: A plot showing the influence ofpH on the fluorescence of a solution ofa Weak
acid witIa p/C. of 7. The measured ftuoreace is assumed to be due to the conjugate
acid. The fluorescence intensity of a strongly acid solution was assigned a value of 100 and
the intensities at other values o( pH
were calculated relative to this.
setting that t
ransmitted fluorescence that was specifically due to the un-
ionized form of the acid. Then
5, = (HA)
- (4.15)
Where h is a combination of all
the
constants of Eq. (4.-6). Hpwever, since
(4.16)
K. + (H)
where c is the Sloi
chiometnc concentration of weak acid, K. is the disso-
ciation constant of the weak acid, and (H) is the concentration of solvated

19. 0
.6
C
6
0
C
4C
a
20
4.5
FACTORS INFLUENCING INTENSITY OF FLUORESCENCE
lBS
protons. Eq. (4.15) becomes
'
cUA('.)
(4.17)
4 .*
An intensity-pH profile such as
that shown in Fig. 4.9
would be expected.
Similarly, if fluorescence which is specific for the conjugate base is detected,-S -
10 12
PH
FIGURE 4.10: A
plot showing the influence of pH on the fluorescence of a solution of a
weak acid with a pK. of 7. The measured tluoescncn is assumed to be duc to the
conjugatebase.
The flu
orescence intensity of a strongly alkaline solution was assigned a value of
ZOO
and the intensities at other values of pH were calculated relatje to this.
then Eq.. (4.6) becomes
K + (H)
(4.18)
and the pH profile illustrated in Fig. 4.10 would be expected.
more complex pH effect can be observed 'ith some compounds and
IS
due to the acid strength of a molecule in the excited state being different
from the acid
s
trength in the ground State. This difference has been shown to

20. 50
sq
-z 40
a
30
.2 20
0
C
w
tO
186 nxoRoltTlv [Cu. 4)
he quite marked for a number or compounds. The phenomenon is known as
excited-siare dissociation. It was first studied by Forster" and was recently
discussed by Ellis.'6 The occurrence is best illustrated by example, and
Fii. 4.11 illustrates such a case. Here, the relative intensity of fluorescence
for solutions of 2-naphthol was plotted as a function of pH. This phenolic
2 3 4 5 6 7 8 9 iC ii 12 13
DH
FIGURE 4. LI A plot showing the iiifince of pH on the intensity of fluorescence of solu-
tions of 2-naphthol. An emission filter which passed wavelengths loner than 415 mn was
employed. The measured fluorescence was, therefore, due to excited-state anions. Re.
printed from Ref. 16.
p.
261, by courtesy of the Journal of chemical EducaFion.
compound has a PKa of 9.5. Both ionized and un-ionized species fluoresce,
but the fluorescence peak for she ionized form is at 429 mu, while that for the
un-ionized species is at 359 m, The fluorcscencc measurement plotted"in
Fig. 4.8 was, therefore, due to the 2-naphthoate anion If the degree of
ionization of the acid in the eround state was the only property influenced by
a change in the concentration of hydrogen ion, then no observable fluores-
ccnce would be expected until a pH was achieved where a detectable degree of
ionization occurred, i.e., a pH of 8.5 (pK0 - 1). That fluorescence was

21. 4.5 FACTORS INFLUENCING INTE\SXTY OF FLUORESCENCE 187
observed in the region of 2 to 8.5 is indicative ofexccd.tatc dissociation. A
possible mechanism. COflSISSCnI with the data, is given by:

ROH
Ro-
\
4 /

ROH + hr. .- ROH RO + H

ROH +
RO- + hl'.
The processes are
I. Absorption of radiation by the un-ionized form
2.Fluorescence of the un-ionized form
3.Radiationkss dissipation of energy
4.Excited-state dissociation to produce a proton and an anion in the
excited state
5.Fluorescence of the anion
6.Radiationless dissipation of energy
The fluorescence at 429 my exhibited by 2-naphthol in the pH range of 2 to
8.5
means that excited-state anions did exist in this range and must have re-
sulted from excited-state dissociation of the molecular species. The first
inflection point in the pH profile corresponds to the cxcited-state dissociation
constant, which is approximately 2.8 for 2-naphthol. The second inflection
point, of course, corresponds to the ground-state dissociation constant.
Practical consequences of the pH effects on fluorescence are the same as
those encountered in spectrophotometiy. Thus, with certain compounds,
pH must be considered as an experimental variable that is important to con-
trol. In some instances, it may be possible to utilize the influence of pH to
minimize interferences and to conduct differential fluoromets-jcdetermina-
tions by appropriate pH adjustment. Additionally, fluorornetry might offer
a convenient approach to the determination of acidity constants of some
compounds.
0. TEMPERATURE
The quantum efficiency of fluorescence is found to decrease with an increase
Iu s he
effective are radiationlcss processes in dissloating excitation energy. It is
Icit
sign I fi-
cantly influenced by temperature changcs.--Thc effect is most probably due to
an increase in the thermal motion of molecules at hi
gher temperatures. The
increased motion favors the probability of intermolecular collision and
subsequent energy loss. In general, a rise in temperature of 1°C results in a
decrease in the intensity ofluoresccncc of about I . For some compounds,
the sensitivity of fluorescence to temperature is even more pronounced.

22. 88 FLUOROMFIRY [CH. 4]
Because of the temperature effect. a reasonable degree of temperature control
is necessary in fluorometric methods. It is recommended, for example. that
analytical samples be equilibrated to the same temperature before measure-
ments are made. Similarly, readings should be taken within a reasonable
period of time to preclude the heating of a sample contained in the sample
holder of an instrument. .. -
\%ITHER FACTORS
I. Degradation of Sample
The stability of a compound in an analytical sample is of concern in fluor9-
metric methods, as it is in other methods of analysis. Here, in addition to
autoxidative and solvolytic degradative routes, the possibility of photolytic
degradation should always be anticipated. Many compounds arc subject to
light-catalyzed degradations and rearrangements. In fluorometric measure-
ments the intensity of light used and the amount of sample irradiated can be
sufficiently large to cause measurable loss of material during the time required
for measurement. The error introduced by photodecomposition can be
reduced by decreasing the intensity of exciting light and by making the meas-
urement in as short a time period as possible.
2. Solvent Effects
The medium in which a potentially fluorescing material is dissolved can
influence the intensity and characteristics of fluoresced light- As previously
mentioned, impurities in solrents can contribute artifactual fluorescence.
In addition, Raman scattering of the exciting light by the solvent might be
erroneously measured as being due to the fluorescence of a sample. This
effect is usually not of practical significance except when measurements are
made on very dilute solutions Other effects can be encountered such as the
quenching of fluorescence by the solvent or by substances such as oxygen
dissolved in the solvent. The spectral characteristics of fluoresced light can
vary from solvent to solvent due to polarization effects and hydrogen bond:
ing. Since these effects cannot be quantitatively predicted, the solvent cannot
be changed at random in fluorometric methods of analysis.
\V%"
0MSo OF FLUOROMETRY WITH
SPECTROPHOTOMETRY
A SENSITIVITY
Fluorometry is significantly more sensitive as an analytical tool than is
spectrophotometry. In spectrophotometry the intensity of light transmitted

23. 46 COMPARISONS OF FIXOROMETRY WITH SPEc -TROPHOTOMr-rRY 189
by a sample is measured and compared to that transmitted by a blank. The
lower limit of detectability is determined by the smallest concentration that
will yield a detectable intensity difference between sample and blank. As the
transmittance of a sample approaches I, small errors .tnade in measuring the
difference between the two intensities result in large errors in calculated
concentration. The lowest limit of concentration that can be detected with
accuracy is. for all practical purppses, established by the molar absorptivity
of the compound under investigation. A fluorometcr measures directly the
intensity of fluoresced light. Moreover, a decrease in the concentration of
fluorescing species can be compensated for by an increase in the intensity of
exciting light and/or an increase in the sensitivity of the detector. It is the
latter variable that most significantly contributes to the sensitivity of the
method. The directly measured intensity can be amplified more readily and
accurately than the intensity difference measured in spectrophotometry. The
lower limit of concentration here is, therefore, established by characteristics
of the instrument and not usually by characteristics of the fluorescing species.
In practice, it is the level of inherent "noise" of the instrument relative to the
signal caused by sample fluorescence that dictates the lower limit of detect-
ability: It has been calculated 17
that fluorescence measurements can offer
sensitivity increases of as high as 103-104 over absorbance measurements.
B. SPECIFICITY
A fluoroinetric assay can offer a degree of specificity that might not be
attainable with a corresponding spectrophotometric technique. Not all
compounds which absorb ultraviolet and visible light fluoresce, and so a
potentially interfering compound which absorbs light will not necessarily be
a source of interference in a fiuorometric method. In addition, the analyst
has the ability to vary the wavelengths of both exciting and fluorescing light
and to choose a combination of wavelengths which 'will maximize the meas-
ured fluorescence from the compound in question and minimize contribu-
tions from interfering substances. It should be noted in this respect that
measurements aced not be made at wavelengths corresponding to maxima or
minima of the spectra. The equations relating fluorescence intensity to con-
centration hold for any region of the spectrum.
C. EXPERIMENTAL VARIABLES
It is obvious from the previous discussions that there arc a larger number
of experimental variables that must b controlled in fluorometric methods
of analysis than in corresponding spectrophotometric methods. For example.
temperature and the intensity of incident light must be maintained reasonably
constant in a fluorornetric method, but need not be rigidly controUcd in 3
spcctrophotomctrjc procedure. Extraneous solutes can markedly aitcct the

24. 190 FLUOROMETRY
[cii. 4]
intensity of fluorescence by quenching effects, whereas in spcctrophotonictry
it is unusual to encounter a system in which the absorbance of a compound
is significantly altered by the presence of other solutes. In addition, the
influence of pH on fluorescence can be much more complex than on absor-
bance and might necessitate closer control of pH in fluorometric procedures
than in spectrophotometric assays.
4.7 APPLICATION OF FLUOROMETRY TO
PHARMACEUTICAL ANALYSIS
An examination of official compendia might lead to the impression that
fluorometry has limited application in the analysis of drugs. There are, for
example, only thirteen monographs in USP XVII, BP (1963), and NF Zn
that specify the utilization-of fiuorometiy as an assay tool and these are
concerned with the determination of only two componnds, riboflavin and
thiamine. However, -a survey of the literature reveals that ftuoromeuy has
enjoyed widespread use in the analysis of drugs in system other than dosage
forms. The sensitivity of the method has resulted in its application in a host of
pharmacological, biochemical, toxicological, phannacokinetic, and bio-
pharmaceutical studies for the analysis of small amounts of drugs in biological
fluids and tissues. It is not practi-.a1 nor necessary to review all such applica-
tions in this chapter. These have been discussed in some detail by Uden-
friend, Phillips and Elevitch,1' and Williams and Bridges.' More recent
reviews are found in the annual Analytical Review editions of the journal
Analytical Chemistry. In 1965, Wimer et al.0 reviewed the literature appear-
ing in the period 1962-1964 which dealt with the analysis of pharmaceuticals.
In 1966, White and WCiSS1&' reviewed publications pertinent to the subject
of fluorometric analysis, including those relating to pharmaceutical analysis
and which appeared in the period 1963-1965. It is illuminating and illustrative
of the wide applicability of fluoromy to drug analysis to note from these
reviews that in the period 1962-1964, pi Tit apared which described
fluorometric procedures for the analytical determination of the following
drugs or classes of drugs: adrenaline, aldoseerone, androsterone. antic.
histamines, atropine, barbituies, ch1i'promazine, chloi'prothine, codeine.
dpyridamole. metine, ergot alkaloids, estradiol, estriol. esterone, ethinyl
estradiol, fursemide, imipramine, isoniazd, niephenesin, mescaline, mor-
phine, narcotine, panthenol, papaverinc, quinidine, quinine, reserpine,
riboflavin, sa licyiates, streptomycin, sulfonamides, testosterone, tetracyclines,
thebaine, tubocurarjne. yimbine.
A limited number of examples are discussed in the following section to
more specifically illustrate the applicability and utility of fluorometry as an
analytical tool. These examples will-also demonstrate that in most instances
th,
e actual determination of the intensity of fluorescence of a sample prep-
aration is the terminal step in a series of operations which demand that the

25. 4.7
APPUCATtON OF FLUOROMETRY TO PHARMACEUTICAL ANALYSIS 191
analyst be cognizant of a variety of scparatory and chromatographic tech-
niques, and the chemical and physical properties of the constituents of a
sample.
Amlnoc'rlfle. 'A simple procedure for the determination of aminocnne
(N) in drug preparations was recently described by Roberts° and illustrates
a relatively direct fluorometric assay method. Here, the aminocrme was
extracted with chloroform from a4asic solution; the chloroform was evap-
orated and the residue was dissolved in acidic ethanol. The fluorescence of the
resulting solution was determined using an excitation filter having maximum
transmittance at 365 mu and an emission filter which transmitted light of
wavelengths greater than 415 mz. The concentration of aminocrine was
determined by comparing the fluorescence of the sample preparation to that
of a standard preparation. The method was applied to a variety of amino-
cz-ine.containing dosage forms, including suppositories, creams, ointments.
jellies, and tablets. Other constituents of the dosage forms did not interfere
with the determination. The method appears to be sensitive, as evidenced by
the recommended use of a standard preparation having a concentration of
1 JAgInIl.
Phenothiazlnes. Mellinger and Keeler' published an interesting study
on the fluoresceiice characteristics of phenothiazine drugs. Thioridazinc
(XI) is representative of the compounds studied. All of the phenothia-
©t05—cH.
cHH
CJNCHI
XI)
zines exhibited similar excitation and emission spectra. The emission spec-
trum of a solution prepared in 0.2 N sulfuric acid was characterized by a
single peak in the range of 450 to 475 The excitation spectrum was
found to have two peaks, one at apprOimately 250 mu and the other in the
range of 300 to 325 m. Addition of potassium permanganate to such solu-
tions resulted in marked changes in spectral characteristic.s. The wavelength
of maximum emission shifted to much lower wavelengths and the intensity of
fluorescence at this maximum was much ereatcr (15-20 times) than that

26. 192 FLUOROMETRY [CH. 41
• exhibited by untreated drug at its maximum. In addition, the excitation
spectrum, after permanganate treatment, exhibited four peaks. Evidence
was obtained to indicate that oxidation of the phenothiazine to the corre-
sponding sulfoxide was responsible for this behavior. The spectra of pheno-
thiazine-like compounds such as prothipendyl (XII), chiorprothixene (XIII),
and imipramine (XIV) were affected by permanganate treatment, but the
am
©o..
effects were qualitatively much different than those observed with pheno-
thiazines.
Pretreatment of phenothiazines with potassium permanganate permitted
quantitative fluorometric analysis of simple solutions in concentrations as
low as 0.01 to 0.05 4ug/ml. The sensitivity was somewhat less for samples
obtained from biological sources 'but was ,signiflcantly greater than that
afforded by other techniques. For example, thioridazine in urine could be
detected fluorometrically at a level as low as 0.8 #g/ml. In contrast, a c-
centration of 4 pg/mI was required for. spectrophotometric detection, while
6 pg/mI was necessary for detection byacolorimetric method which was based
on the treatment of a phenothiazme with concentrated sulfuric acid.
Sahcylates. In 1948, Saltzman22 described a fluorometric method for the
estimation of salicylates in blood. It was based on the observation that
salicylates, in alkahac medium, exhibit a blue fluorescence. The procedurç
involved precipitation of proteins from a sample with dilute tugstic acid and
treatment with strong alkali. Alternatively, a sample of plasma was acidified
and extracted with ethylene dichlordc. The salicylate was then back-
extracted into strong alkali. The fluorescence of the resulting solution was
me.sured using a 370-mu excitation filter and a 460-mp emission filter.
Chirigos and Udcnfriend23 studied the fluorescence characteristics of sali-
cylate in more detail and determined from spectrofluorometric studies that

27. 4.7 APPLICATION OF FLUOROMETRY TO t'IARMACL1.TCAL ANALYSIS 193
the excitation maximum was at 310 niju and the emission maximum was at
400 mu.' Salicylate content of biological tissue was determined by extracting a
sample with ether, back-extracting into a borate buffer, and tluoromctric
examination of the borate solution. More recently, Lance and Bell2' used
fluoromctry coupled with paper chrornatpgraphy as the basis for a micro-
method for the determination ofacctylsalicylic acid and salicylic acid in blood
sample. Here, a small sample of blood (tOO p1) as extracted with ethylene
dichloride.' Aliquots of the ethylene dichloride extract ere spotted on paper
strips and the strips were developed by the ascending technique using 0.75
nitric acid as the solvent system. Segments containing the aspirin and the
salicylic acid were cut from the strip and eluted with 5 Al sodium hydroxide.
The fluorescence of solutions prepared in this manner were determined using
as a blank a solution prepared by treating another portion of the chromato-
graph strip with alkali. Standard curves were prepared for the two com-
pounds by subjecting blood samples containing known amounts of the drugs
to the procedure and relating observed fluorescence to concentration. The
method was applied to an investigation of aspirin and salicylic acid blood
levels after the oral administration of aspirin tablets.
HomovnIlIIc AcId. Homovanillic acid (3-methoxy-4-hydroxylphenylacet ic
acid, XV) is derived from the metabolism of 3,4-dihydroxyphenylalartine and
3 ,4-dihydroxyphenylethylaminc. The urinary excretion of homovanill ic
acid (HVA) is elevated in patients with ocuroblastoma and pheochromo.
cytoma and interest has been expressed in utilizing urinary levels of HVA as a
diagnostic tool. Anden et al. and Sharman26. discovered independently that
H'VA reacted with oxidizing agents such as potassium ferricyanide and
ferric chloride, in alkaline medium, to yield a highly fluorescent solution.
Although the product responsible for the fluorescence was not identified,
both groups developed fluorometric procedures for the determination of
HVA in biological tissues. The procedure of Anden et al. was modified by
Sato,"", who incorporated an ion-exchange treatment and solvent extraction
into the procedure to isolate the HVA. which was then treated with ammonia
and potassium ferricyanide. The fluorescence or the resulting solution
was then determined. Corrodi and Werdinius more recently studied the
procedure in more detail and determined that the compound which was re-
sponsible for the fluorescence was 2,2-dihydroxy-3.3'-di met hoxy-biphcnvl-
5,5'-diacetie acid (XVI). This compound exhibited an excitation maximum
COOH COOH
CH,COOH CH, CH,
OCH, CH3O y

0CM,
01! OH OH
(XV) (XVII
at 315 mu and an emission maximum at 425 mp.

28. 194 ILUOROM(TRY ' (cii. 41
An interesting application of this oxidative transformation of HVA was
suggested by Guilbault et al.' They found that the conversion of XV) to
(XVI) could be accomplished enzymatically in a hydrogen peroxide3peroxi-
dasc system. They frmuIated solutions containivg HVA', hydrogen per-
oxide, and peroxidase and measured the rate of change of fluorescence. At a
constant concentration of HVA. this rate was found to be directly propor-
ttonal to the peroxidase concentration and to the concentration of hydrogen
peroxide. They recommended this approach for the analysis of oxidative
enzymes and of hydrogen peroxide. They reported that as little as 10-"
mole/liter of peroxide and 10 unit/ml of peroxidzise are determinable by this
method.
Cardiac Glycosides. Jakovijvic reviewed the many methods which have
been proposed for the determination of cardiac glycosides. He noted that a
number of the reported methods are fluorometric and are based on the gen-
eration of fluorophors by the action of dehydrating agents on the steroid
moiety of the glycoside. He proposed a new reagent for this purpose, a
mixture of acetic anhydride, acetyl chloride, and trifluoroacctic acid. He
described the use of this reagent and a procedure for the simultaneous deter-
mination of digitoxin (XVII) and digoxin (XVIII) in mixtures by fluoromctry.
The method is an interesting illustration of how assay specificity can be
attained by utilizing a knowledge of both excitation and emission charac-
teristics of fluorescing compounds.
.xvm (XVIII)
Digoin differs from digitoxin by one hydroxyl group at position 12.
However, when the two compounds were treated under anhydrous conditions
with the dehydrating agent, they yielded products which had significantly
different fluorescence characteristics. The fluorophor generated from digi-
toxin exhibited a single excitation peak at 470 my and a single emission peak
at 500 mhz. The di-oxin fluorophor exhibited two excitation peaks at 35
and 470 ma. Excitation at 345 mu resulted in an emission peak at 435 m1,
while excitation at 470 mu gave an emission peak at 500 m. Under the
latter conditions, the fluorescence was approximately 30% of that obtained
with digitoxin When the digitoxin preparation as examined at an excitation
wavelength of 345 mi and an emission wavelength of 435 nip, no fluorescence
was observed. The author suggested that the treatment of digitoxin resulted

29. I
47 APPLICATION OF FLUOROMETPX TO PhARMACEUTICAL ANALYSIS 195
in the formation of a substituted 3.4-benzpyrcne, while digoxin yielded a
mixture containing compounds related to 3.4-benzpyrcne and chrysinc. The
differences in spectral characteristics permitted the simultaneous dctcrrnina-
tion of digitoxin and digoxin in samples prepared from digitalis leaf and
--.4jgita1is tincture. Here, the glycosides were isolated by extraction from a
sarnfrad with the dehydrating agent for 30 min at 45°C, and diluted
with dichioromethane. The fluorescence of the resulting solution was deter-
mined with a fluorometer using two different filter combinations. One com-
bination was equivalent to an excitation wavelength of 470 mu and an
emission wavelength of 500 m. The measured fluorescence under these
conditions resulted from both digitoxin and digoxirt. The other combination
was equivalent to an excitation wavelength of 345 mp and an emission wave-
length of 435 m. fluorescence here was due to digoxin. The later reading
could be used to calculate, by comparison to a standard, the concentration of
digoxin. A knowledge of this concentration was then used to calculate the
fluorescence due to digoxin under the excitation and emission conditions of
the first filter combination and to obtain a corrected fluorescence which
reflected the digitoxin concentration. The corrected fluorescence was then
used to calculate the digitoxin concentration by comparison to that exhibited
by a standard. The method was also applied to the determination of di gi toxin
in tablets and ampoules.
Reserpine.'The reaction of nitrous acid with reserpine (XIX) to yield a
yellow fluorescent pigment was described by Szalkowski and Mader and
cxIx
forms the basis fora widely used coidrimetric method for the determination of
reserpinc in pharmaceutical preparations. Reserpine has been used as an ad-
ditive to poultry feeds, where it is found in concentrations as low as 0.0001 %.
The low levels encountered in such systems could not be satisfactorily deter-
-mined by the colorimetric method of analysis. Mader et al.32-33 utilized the
fluorescence characteristics of nitrous acid-treated rcscrpine to obtain the -
desired sensitivity. treatment of rescrpine with nitrous acid was found to
result in-the formation of a product whih possessed an excitation maximum
at 390 mu and an emission maximum at 510 rnu. The intensityoffluorcsccrtce
was linearly related to concentration over a wide range. The reserpinc in a
feed sample was isolated by a series of extractions and was eventually ob-
tained as n assay preparation in chloroform-methanol. An aliquot of this

30. 96 FLt.JOROMETRy (CH. 41
solution was treated with sodium nitrite and the mixture was acidified with
hydrochloric acid. Alter an appropriate reaction time, sulfamic 1acid was
added to consume the excess nitric acid and the fluorescence of thsolution
was determined. A blank was prepared by treating an aliquot of the assay
preparation in a similar manncr,but with the omission of the sodium nitrite.
The concentration of reserpinc in the assay preparation was calculated by
comparison of the measured fluorescence to that found with a standard
preparation which was obtained by carrying a known amount of reserpinc
through the extraction and reaction procedures.
Haycock et al.1 recently reported the results of study in which the kinetics
and mechanisms of the nitrous acid-induced fluorescence of reserpine were
investigated. They presented evidence to show that the fluorescence was due
to the formation of 3-dehydrorcser-pine. The kinetics of the reaction indicated
that protonated reserpine initially reacted with nitrous acid to form an inter-
mediate complex, which then underwent an acid-catalyzed reaction to form
3-dchydroreserpie
Vitamins. Udenfriend33reviewcd the many fluorometric methods that have
been used for the determination ef vitamins. He discussed, in some detail,
procedures for vitamin A, thiamine, riboflavin and related flavins, nicotin-
amide, pyridoxine and related compounds, ascorbic acid, vitamin D, folic
acid, p-arninobenzoic acid, cyanocobalarnin, tocopherols, and vitamin K.
The fluorometric determinations of thiamine and riboflavin are of special
interest in that they serve as rather classic examples of the application of
fluorometry to pharmaceutical analysis.
Thiamine (XX) possesses little native fluorescence, but is readily oxidized
to thiochromc (XXI), which is highly fluorescent. Thiochrome has been
CH 3_^^r tCH10H
CHI_rJ.?4)(I°I
(xx)
reported to tave an excitation maxima at 365 mz and an emission maximum
of 450 m.34 The procedure described in the seventeenth revision of the
USP serves as an example of the fluórometric assay. An aliquot of a sample
solution of the vitamin is treated with -an oxidizing reagent (an alkaline
solution of potassium fcrricyanide). The thiochromc which is formed is
extracted into isobutanol and the fluorescence of the resulting solution is
determined. The fluorescence is corrected by use of a blank and is compared
to that of a standard preparation.
(Riboflavin has a characteristic pronounced native fluorescence and can be
assayed by direct fluorometric cxarnination of a sample solution-'. As will
be discussed in the practical section, tluoromcrric assays for riboflavin usually
employ an internal standard and an intcrnal blank.

31. 4.8 PRACTICAL SECTION 197
4.8 PRACTICAL SECTION
A. GENERAL
The procedures used in fluorometric assays are, in most instances, quite
similar to those encountered in colorimetry and spectrophotometry. Treat-
ment of a sample frequently involves dilution, extraction and/or chromato-
graphic separation, chemical reaction, and finally the determination of the
intensity of fluorescence of the final assay preparation using a suitable fluorom-
eter fitted with appropriate filters. The intensity is "read cut" in arbitrary
units and, after correction for the fluorescence contributions for the blank,
can be used to estimate the concentration of fluorescing substances. For'*
nonlinear intensity-concentration profile, a standard curve must be used for
this purpose. if the dilution is such that intensity is directly proportional to
concentration, the concentration of the assay preparation can be obtained by
comparing the intensity reading to that of a single, standard preparation.
In such a case, concentration is calculated by a formula, familiar from spec-
trophotometric assays:
C CI,FJF, (4.19)
where C is the concentration of the assay preparation, C. is the concentra-
tion of the standard preparation, F is the fluorometcr reading, corrected
for blank, obtained with the assay preparation, and F, is the fluorometer
reading, corrected for blank, obtained with the standard preparation. The
!rtferna:ionalPlzonnacopoea3' cautions that the ratio FJF should not be less
than 0.04 and not more than 2.50 because of the limited concentration range
within which fluorescence is proportional to concentration.
Frequently, internal standards are prescribed in fluorornetric procedures.
Here, a known quantity of pure material is added to the assay preparation
to compensate for quenching effects which might be introduced during the
work-up of a sample. The USP XVII assay for riboflavin3s serves as an ex-
ample. This assay specifies the treatment of 10 ml of an assay preparation
with I ml of water and 2 ml of reagents. The fluorescence of the resulting
solution is measured and designated I,. Another JO ml of assay preparation
is treated with I ml of a standard preparation containing 0.001 mg of ribo-
flavin per ml and 2 ml of reagents. The fluorescence of this solution is
measured and designated I,. The concentration of vitamin in the assay prep-
arationin milligrams per milliliter is calculated by the formula:
C
- '
x 1113 0.001 x 13/10 = - x 0.0001 (4.20)
where 'b is the fluorescence reading obtained with a blank.
Various procedures are used to obtain blank readings in fluorometry.
Conventional blanks are sometimes specified and arc prepared by substituting

32. 198 FLUOR0Mfly
(CH.
41
in the final step of an assay a volume of water or buffer for the required vol..
umc of assay preparation. When the possibility exists of fluorescent materials
being introduced to the assay preparation by the system containing the
compound of analytical interest, a more realistic blank
i5 usually recom-
mended. For example, plasma and urine blanks are prepared by carrying a
volume of drug-free plasma or urine through the complete procedure. Internal
blanks are frequently employed. Here the assay preparation is used as a
blank after specifically eliminating, through chemical reaction, the fluores-
cence due to the drug. In the riboflavin assay, for example, a few crystals of
sodium hydrosulfite are added to the cuvette immediately after the fluores-
cence intensity of a sample is measured. The hydrosulfite rapidly and specifi-
cally converts riboflavin to the nonfluorescent, reduced form. The
fluorescence of the resulting solution is measured and is used as a blank
reading to correct for fluorescence arising from sources other than riboflavin.
In other instances, blanks are prepared by omitting a reagent necessary for the
generation of a fluorescing species. For example, the official assay for
thiamine° is based on the oxidation of nonfluoresccnt thiamine to strongly
fluorescent thiochrome. The oxidizing agent employed is alkaline potassium
ferricyanide solution. Blanks
for both assay and standard preparations are
prepared by submitting samples to the full procedure, but with the substitution
of a volume of
sodium hydroxide solution for the volume of oxidizing reagent
which is normally used.
Numerous instrumental variables such as the intensity
of exciting light,
detector response, signal amplification, etc., influence the measured intensity
of
fluorescence. Aging of a light source and fatigue of a detector could, for
example, result in nonreproducibility of results and assay error. It is impor-
tant, therefore, to periodically check the sensitivity of a fluorometer and to
adjust it to constant sensitivity during the course of assay measurements. A
solution of a stable, strongly fluorescing substance is used for this purpose
and is known as a comparison standard. The standard chosen should have
fluorescence characteristics similar to those of the compound being assayed
and, in fact, if that compound is sufficiently stable, no other comparison
standard is needed. A solution of quinine sulfate is frequently recommended
as a comparison standard. Its use is illustrated by the USP assay for thiamine.
A solution of quinine sulfate in 0.IN sulfuric acid at a concentration of 0.25
P9/ml is recommended since "this solution fluoresces to approximately the
same degree as the thiocl-irome obtained from I pg of thiamine hydrochloride
and is used to correct the fluoromcrcr at frequent intervals for variations in
sensitivity from reading to reading within an assay."
B. LABORATORY PROJECTS IN FLUOROMETRY
The following projects arc offered as guides for possible laboratory exercises
illustrating some principles and applications of fluorometr
y. Since different

33. 4.8 PRACTICAL SECTION 199
makes of fluoronictcrs differ in sensitivity, ranges of sensitivity, and the
manner by which ranges of sensitivity are selected and adjusted. exact
experimental details cannot be presented. The student should initially become
familiar with the operational characteristics of the fluorometer available for
his use by studying the instructional and descriptive literature supplied with
the instrument and by appropriate laboratory demonstration.
I. Intensity of Fluorescence of Riboflavin as a Function of Concentration
Prepare a stock solution of riboflavin at a concentration of about I pg/mi
(USP XVII or NF Xli may be consulted for directions for preparing this
solution). Prepare dilutions of the stock solution to obtain the following
concentrations: 0.02, 0.04, 0.06, 0.08, and 0.1 ag/mI. Determine the relative
intensity of fluorescence of each solution with a suitable fluorometer. An
appropriate primary (excitation) filter is one that peaks at 360 mp, while the
secondary filter (emission) should pass wavelengths greater than 510 mp.
Present the results in the form of a graph in which fluorometer reading is
plotted as a function of concentration. Repeat with solutions ranging in
concentration from 0.002 to 0.01 pg/mI. -
2 The Influence of pH on the Fluorescence intensity of Riboflavin
Prepare buffered solutions of riboflavin ranging in pH from 2 to 11. All
solutions should contain the same concentration of the vitamin, which should
be such that the solution buffered to approximately pH 7 gives a reading of
from 50 to Boy. of fuji scale of the fluorometer. Determine the fluorescence
of each solution and plot the fluorometer reading as a function of pH.
3.The Influence of Quenching Agents on the Fluorescence
Intensity of Riboflavin
Design and conduct an experiment to demonstrate the influence of potas-
sium iodide concentration on the intensity of fluorescence of riboflavin.
Plot the results in a manner suggested by Eq. (4.11). Repeat using caffeine
as a quenching agent.
4.The intensity of Fluorescence of Salicylic Acid as a Function
of Concentration
Prepare solutions of salicylic acid in 0.1 N sodium hydroxide to cover a
rani!e of concentrations of from 5 to 100 og/mI. Determine the relative
intensity of fluorcsccncc of each solution and plot fluorometer reading as a
function of concentration. The 360-m,uprimary filter may also be used in this
instance, but a secondary filter transmitting wavelengths greater than 455 mp
should be selected.
S. Assay of Riboflavin Injection
Determine the potency of a sample of riboflavin injection by empliiflg
the riboflavin-assay procedure described in USP XVII or NF Xli.

34. 200 FLLJOROMETRY
(CH. 4)
6.
Assay of Thiamine Hydrochloride Tablets
Employ the thiamine assay procedure described
in (iSP XVII od NF XII
to determine whether or not a sample of thiamine hydrochloride table
the label claim.
Meet
-
7.
The Determination of 9-Amjnoacrpdine (Aminocrine) In
Pharmaceutical Products
Obtain a sample of a pharmaceutical preparation containing aminocrjne.
Fluoromctrically determine the aminocrine content
by the method proposed
by Robcrts.20
8.Exejcedstate Dissociation of 2-Naphthol
Conduct the laboratory experiment described by Ellist6 to demonstrate
the excited-state dissociation of 2-naphthol.
9.Fluoromei.y In Biopharmaceutical Studies
Determine the physiological availability of riboflavin from a coated tablet
using the method of Chapman et aL
40 In this procedures the amount of ribo-
flavin excreted in the urine following oral ingestion of a coated vitamin
tablet is determined fluorometrically and compared with the amount excreted
alter the ing
estion of a rapidly dissolving uncoated tablet.
PROBLEMS
P4.1. The molar absorptivity of riboflavin (molecular weight — 376.36) in aqueous
solution at 360 mu is approximately
7500. Fluorometric examination, using
an excitation wavelength of 360 mp, of a solution of the vitamin containing
0.010 pgJml yielded a fluorescence intensity of 1.0 unit. Calculate the theoreti-
cally expected intensity of fluorescence for a solution containing 1.0 pg/mI.
Assume chat the equivalent of I-cm cell was used.
P4.2. The influence Of PH on the intensity of fluorescence of dilute solution of
weak arid was investigated. Solutions were prepared which varied only in
the concentration of hydrogen ion. The fluorescence of each solution was
determined with a fluorometer and the following results were obtained:
Fluorcscencc
pH
(arbitrary units)
4.0
80.0
5.0
80.0
6.0
79.5
7.0
745
7.5
65.6
8.5
34.4
9.0
25.4
10.0
206
11.0
200
12.0
200

35. REFERENCES 201
Show that the ratio of the concentration of ionized acid to that of un-ionized
acid at any p1-t is given by (80 - - 20).
where F is the observedfluorescence at that pH. Plot
thc
logarithm of the ratio as a function of pH
and determine the pK of the acid from the plot.
P4.3. A standard preparation of riboflavin was prepared by the following procedure
Exactly 48.5
mg of USP riboflavin reference standard was dissolved in suffi-
cient water to make 500 ml. One ml of the resulting solution was diluted to
200 ml with water.
An assay preparation was prepared in the following manner: The riboflavin
from 10 riboflavin tablets was dissolved in
sufficientwater to make I liter,
One ml of this solution was diluted to I liter with water.
A mixture of
10 ml of assay preparation and I ml of standard preparation
yielded a fluorescence reading of 74.0. Ten ml of the assay preparation and
I ml of water gave a fluorescence reading of 40.0. The latter mixture was
treated with 20 mg of sodium hydrosulfite and the fluorescence was again
d
etermined. The reading was found to be 2.0. Calculate the quantity in mg
or riboflavin in each tablet.
P4.4. A n
onfluoresmnt complexing agent A is known to form a nonfluorescent'
complex with a fluorescent compound B. A I x JO"
M solution of .B was
placed in a fluorometer and the meter was adjusted to read 100.0. The
fluorescence of a solution which
was I x 10'M with respect to B andI x 10— M
with respect to A was then determined and
was found to be
20.0. Calculate the stability Constant of the complex.
P4.5. Exactly 20 mg of liSp
thiamine hydrochloride reference standard was
dissolved in sufficient water to make I liter. One ml ofthe resulting solution
was diluted to 100 ml with water. Five ml of this solution was treated with
3 ml of oxidizing agent and the thiochromc which was formed was extracted
into 20 ml of i
sobutanol. The fluorescence of the isobutanol phase was
determined to be 67.0. A
corresponding blank has a fluorescence of 4.0.
Exactly I ml of a thiamine hydrochloride injection was diluted with
sufficient water to make I liter. Two ml of this Solution was diluted to I liter
with water. Five ml of the final dilution was oxidized and extracted as
previously described, and the
flu
orescence of the isobutanol extract was580. The fluorescence of a
corresponding blank was 8.0. Calculate the
quantity, in milligrams, of thiamine hydrochloride in each milliter of injection.
P4.6. A preparation contains riboflavin thiamine
hydrochloride and a fluorescent
coloring agent. Describe how you would approach the problem of developing
a fluOrometr,c method for the determination of,
ll three ol these components.
REFERENCES
I. R. T. Williams and J. W. Bridges.].
C/in. Path., 17, 371 (1964),2.1). S. McClure J. Chem. ph I,_
17. 905 (1949)
3.
E. L. Wchry and L. B.
Rogers in Fluorescence and Phorphorrsct,, Analrcis (D. M.Hcrcujes. ed.). Wile
y (Inlersciencc) New York.
1966, Chap. 34.K. W.
Kohn Anal Chrrn. 33, 862 (1961)
5.D. Matungly. f. C/in. Path., 15, 374 (1962).
6.3 W
ilurr. J.
Thai, Chem,, 167, 151 (1947
7.H, L. Price and M. L. Price,
J. Lail. C/rn. Med, 50, 7,j (1957)

36. 202 FLC0R04ETRY [CH. 41
8.P. L. Lott. 1. Chem. Educ., 4l, A327, A421 (1964). -
9.5, Udenfriend. in Fluorescence .4jsay in Biology and Medicine. Academic Pre-n. New
York, 1962, Chap. 3.
10.F. Kavanagh. Ind. &tg. Chem. Anal. Ed., 13, 108 (1941). . .
II. H. Braunsberg and S. B. Osborn. Anal. Chin,.Acia. 6, 84 (1952). -
12.D. M. Hercules, Anal. Chem., 38, 29A (1966).
13.D. M. Hercules, in Fluorescence and Phosphorescence Analysis (D. M. Hercules, ed.),
Wiley (lncerscicflcc). New York, 1966. Chap. I. --
14.H. Harbury and K. A. Foley, Proc. Nail. Acad. Sci. U.S.. 44. 662 (1958).
15.T. Forster, Nazurwiss., 36, 186 (1949.
16.D. W. Ellis. I. Chem. Educ., 43, 259 (1966).
17.R. E. Phillips and F. R. Elcvitch, in Progress In Clinical Pathology. Vol. 1 (M. Stefanini,
ed.), Grune & Stratton. New York. 1966, Chap. 4. - -
18.D. C. Wimcr, .1.0. Theivagt, and V. E.Papendick. Anal. Chem..185R. (1965).
19.C. E. White and A. Weissler. Anal. Chem., 38, 155R. (1966). -
20.L A. Roberts. I. Assoc. Offic. Age. Chemists, 49, 837 (1966). • .- -. -.
21.T. J. Mellinger and C. E. Keeler, Anal. Chem., 35, 554 (1963). .
22.A. Saltzman, J. Bial. Chem.. 174, 399 (1948).
23.M. A. Chirigos and S. Udenfriend, J. Lab. C/In. Mcd.. 54, 769 (1959). ..,,
24.W. E. Lange and S. A. Bell. J. Pharm. Sc., 55, 386 (1966).
25.N. F. Anden, D. E. Roos. and B. Werdiniut, Life Sc!., 2, 448 (1963).
26.D. F. Sharman, Brit. J. Pharmacol., 20, 204 (1963).
27.T. L. Sato, J. Lab. Clin. Mcd., 66, 517 (1965).
28.H. Corrodi and B. Werdinius, Ada ChemScand., 19. 1854 (1965).
29.G. C. Gtiilbault. D. N. Kramer. and E. Hackley, Anal, Chem., 39, 271 (1967).
30.1. M. Jakovljevic. Anal. Cheni., 35, 1513 (1963).
31.C. R. Szalkoki and W. 1. Mader_ .1. Am. Pharm. Assoc. Sc!. Ed., 45, 613 (1956).
32.W.J. Mader, R. P. Haycock, P. B. Sheth, and R. T. Connally, J. Assoc. Offic. Age.
Chemists, 43, 291 (1960); 44, 13 (1961).
33.W. J. Mades, 0.1. Paparicllo, and P. B. Shcth,J. Assoc. Offic. Age. Chemists, 45,589
(1962). . .
34.R. P. Haycock. P. B. Sheth, T. Higuchi, W. J. Mader, and G. 1. Papariello. I. FhflT.
Sd.. 55, 826 (1966). ..' . . . .
35.S. Udcnfricnd, in Fluorescence Assay In Biology and Medicine, Academic Press. New
York, 1962. Chap. 7. -
36.W. E. Ohnesorge and L. B. Rogers, Anal. Chem.. 28,1017 (1956).
37.international Pharmacopoca. Vol. 11, World Health Organization, 1st ed.. 1955,
Appendix 3, p. 233.
38.United States Pharmacopeia, Mack Printing, Easton, Pa.. 17th rev. ed.. 1965, p. 886.
39.United States Pharmacopela, Mack Printing, Easton. Pa,, 17th rev. ed.. 1965. p. 888.
40.D. G. Chapman. R. Crisaflo. and 1. A. Campbell, I. Am. Pharm. Assoc. Sd. Ed., 43.
297 (1954).
41.C. Bedford, K. J. Child, and E. G. Tomich, Nature. 184, 364 (1959).